Clearing a Stuck Dape Drive

It happens to all of us, some type of error during our backup or cloning and NetWorker leaves a tape stuck in the drive. This tape usually ends up in the "writing idle" or "ready for writing" mode, and there appears that nothing we do can un-mount this drive.

Use the below procedure to correct the situation without bouncing your NetWorker Server.

Step 1 - Open your tape silo and manually dismount the tape. Place this tape in the CAP or just hold on to it until you get the drive working again. Step 1 is optional; you can (usually) get your tape drive working without manually dismounting the tape.

Step 2 - Determine which nsrmmd process is using the drive. There are several methods that can be used to obtain this information:
fuser Method: Enter the following command, fuser device_name
(Example: fuser /dev/rmt/0hbn)
You will receive output similar to the following:
/dev/rmt/0hbn: 1396
The output shows the PID of the nsrmmd process that is associated with the drive.
lsof Method: Enter the following command, lsof device_name (Example: lsof /dev/rmt/0hbn)

You will receive output similar to the following:
nsrmmd 1396 root 11u VCHR 33,468 0xed60000 97625
/dev/rmt/../../devices/pci@1f,0/pci@1/scsi@1,1/st@5,0:hbn

The output shows the PID (1396) of the nsrmmd process that is associated with the drive.
Step 3 - Issue the kill command to kill the nsrmmd process. For example: kill 1396

Step 4 - From the NWAdmin GUI, you should see the nsrmmd automatically restart after you issued the kill command. After the nsrmmd process restarts, your drive should be back in service

Step 5 - Use the nsrjb command to deposit the tape
you manually removed, then inventory the slot where
you place the tape.